A federal ruling issued this Friday in Washington provided new details about the U.S. military operation that ended on January 3 with the capture of the former narcoterrorist dictator Nicolás Maduro.
According to journalist David Alandete from the newspaper ABC, the ruling confirms that Claude, the artificial intelligence system developed by Anthropic, was linked to that mission through Palantir.
The court, however, does not specify what specific task Claude performed during the operation. The ruling does state that an executive from Anthropic questioned the use of technology in a "sensitive military operation abroad".
Furthermore, the ruling itself connects that episode with the mission to capture Maduro and subsequently describes it as a kinetic operation aimed at capturing a "foreign head of state".
Claude
Alandete noted that Claude already had a prior relationship with the U.S. security apparatus. Since 2024, the Department of Defense and the intelligence community had been using Anthropic's models to work with classified information.
Among its applications were document processing, threat assessment summarization, and intelligence analysis.
Anthropic had also developed Claude Gov, a version specifically oriented towards agencies linked to national security. Later, the company allowed other uses related to foreign intelligence analysis, weapon system design, and offensive cyber operations.
In the case of Venezuela, the journalist raises various possibilities regarding the utility that the system may have had during the mission. Claude could have been used to process large volumes of intelligence before and during the assault, analyze surveillance information, locate or track individuals, vehicles, or Venezuelan forces, and synthesize data for military leaders.
The former dictator Nicolás Maduro
It could also have been involved in updating the situation on the ground, identifying threats, and planning movements or extraction. However, these functions are presented as possibilities and the ruling does not establish what specific role was played by the artificial intelligence.
The central point of the ruling is related to the restrictions that Anthropic maintained for certain military uses of Claude. According to Alandete's account, one of the company's executives questioned the activities carried out by a Pentagon contractor, which raised concerns within the Department of Defense.
U.S. authorities suggested that the limitations imposed by Anthropic could cause the system to stop responding as expected during a mission. According to the court, officials feared that the software could cease to function or create a situation capable of putting U.S. military personnel at risk.
After the operation, the Pentagon demanded that Anthropic authorize any legal use of Claude. However, after disputes with the company, the decision was made to remove that AI from the Department of Defense systems and an expansion of the U.S. agency's contractual relationship with OpenAI was carried out.
